Leading Learning

Organizations
The Bold, the Powerful, and the Invisible
by Peter M. Senge
published by Jossey Bass, Inc., a subsidiary of John Wiley & Sons Company, 1996 in
The Leader of the Future
Goldsmith and F. Hesselbein, Eds.
The Peter F. Drucker Foundation

Peter M. Senge is a faculty member at the Massachusetts Institute of Technology

and director of the Center for Organizational Learning at MITs Sloan School of
Management, a consortium of corporations, including Ford Motor Company,
Federal Express, Motorola, AT&T, GS Technologies, Intel, Electronic Data Systems, Harley-Davidson, and Shell Oil, that work together to advance methods
and knowledge for building learning organizations. He is the author of the widely
acclaimed book The Fifth Discipline: The Art and Practice of the Learning
Organization and coauthor of The Fifth Discipline Fieldbook: Strategies
and Tools for Building a Learning Organization.

No significant change will occur unless it is driven from the top. Theres no point
in starting unless the CEO is on board. Nothing will happen without top-management
buy-in.
How many times have we all heard these familiar refrains and simply accepted them
as the way things are? Yet there are good reasons to challenge these hoary truisms.
The evidence of top-management impotence is abundant. Everywhere one hears CEOs
and other top executives talking about the need to transform their organizations, to
overthrow stodgy bureaucratic cultures, to become learning organizations.
Nevertheless, the evidence for successful corporate transformations is meager.
Moreover, the basic assumption that only top management can cause significant change
is deeply disempowering. Why, then, in the age of empowerment, do we accept it so

unquestioningly? Isnt it odd that we should seek to bring about less hierarchical and
authoritarian organizational cultures through recourse to hierarchical authority?

Two Views on Leadership

Why do we cling to the view that only the top can initiate significant change? Is it just
our unwillingness to give up a familiar mental model? Is it the fear of stepping out of
line without the imprimatur of the hierarchy? Perhaps, also, there is an element of selfprotectionthe comfort of being able to hold someone else, namely, top management,
responsible for the lack of effective leadership. There is no doubt that a CEO opposed to
fundamental change can make life difficult for internal innovators, but this hardly
proves that only the CEO can bring about significant change. At the very least, shouldnt
we be suspicious of the knee-jerk tendency of people in organizations to look upward
and expect top management to fix things?
Consider a different view: Little significant change can occur if it is driven from the
top. CEO proclamations and programs rolled out from corporate headquarters are a
good way to undermine deeper changes. Top-management buy-in is a poor
substitute for genuine commitment at many levels in an organization, and in fact, if
management authority is used unwisely, it can make such commitment less rather than
more likely.
When I first came in as CEO, says Phil Carroll of Shell Oil, everyone thought, Phil
will tell us what he wants us to do. But I didnt have a clue, and if I had, it would have
been a disaster. Anyone who thinks the CEO can drive this kind of change is wrong,
says Harley-Davidson CEO Rich Teerlink. There are several reasons leaders like Carroll
and Teerlink come to a more humble view of the power of top management. One is the
cynicism that exists in most of our organizations following years of management fads.
When the CEO preaches about becoming a learning organization, people roll back
their eyes and think to themselves, Here we go again. I wonder what seminar he went
to last weekend. Most corporations have had so many flavor-of-the-month initiatives
from management that people immediately discount any new pronouncement as more
executive cheerleading or, as they say at Harley-Davidson, another fine program.
A second reason is the difference between compliance and commitment. When
genuine commitment is needed, hierarchical authority becomes problematic. It seemed
that every year, someone pressured us to change our promotion review process to
incorporate our values, reflects former Hanover Insurance CEO Bill OBrien. But we
never caved in to this pressure. A value is only a value if it is voluntarily chosen. No
reward system has ever been invented that the people in an organization havent
learned how to game. We didnt just want new behaviors. We wanted new behaviors
for the right reasonsbecause people genuinely believed that openness, localness,
merit, and our other guiding values would really lead to a healthier and more

productive work environment (Moral Formation for Managers: Closing the Gap
Between Intention and Practice, in Character and the Corporation, MIT Center for
Organizational Learning Research Monograph, 1994.) Hierarchical authority, as it has
been used traditionally in Western management, tends to evoke compliance, not foster
commitment. The more strongly hierarchical power is wielded, the more compliance
results. Yet there is no substitute for commitment in bringing about deep change. No
one can force another person to learn if the learning involves deep changes in beliefs and
attitudes and fundamental new ways of thinking and acting.
A third reason is that top-management initiatives often backfire and end up moving
organizations backward, not forward. This can occur in obvious ways; for example, topmanagement downsizings and reorganizations that have the side effect of increasing
fear, distrust, and internal competitiveness reduce cooperation and collaboration,
thereby further undermining economic performance. But it can also occur more subtly,
even in changes explicitly designed to improve learning. For example, the 360-degree
feedback process with which all managers must comply not only reinforces a
compliance mentality; it also lessens the likelihood of people surfacing what Harvards
Chris Argyris, in Good Communication That Blocks Real Learning (Harvard Business
Review, July/Aug. 1994), calls the potentially embarrassing information that might
produce real change. This kind of information will only come into the open when
people have genuine trust, curiosity, and shared responsibility, conditions not usually
fostered by mandated programs. Corporate communication programs rolled out from
the top can actually inhibit learning and communication, according to Argyris.
Organizational surveys and focus groups, for example, by focusing attention on
telling top management what is wrong, can block learning because they do nothing to
encourage individual accountability and tend to reinforce the mind-set that only top
management has the power to fix problems.
The merit of these two alternative views regarding top management and change
depends entirely on the nature of the change one is trying to bring about. By and large,
reorganizations, downsizing, corporatewide cost reduction programs, or reengineering
programs can only be implemented from top-management levels. But such changes do
not change corporate cultures based on fear and defensiveness. They will not unleash
the imagination and passions of people and enhance their ability to form genuinely
shared visions. They do not change the quality of thinking in the organization and the
ability of people to understand interdependency. They do not increase intelligence at the
front lines as people confront increasingly complex and dynamic business
environments. And they will do nothing to foster the trust and skills needed by teams at
all levels if they are to reflect on hidden assumptions and to inquire into the flaws in
reasoning lying behind their own actions.
For almost twenty years, I, and many colleagues, have been working with managers
and teams in developing enhanced learning capabilities: systems thinking, improving

mental models, fostering dialogue, nurturing personal vision, and building shared
visions. For the last four years, a group of us at MIT have begun to develop a consortium
of corporations to advance the theory and method underlying this work and to
demonstrate what is possible when organizations get serious and when people work
together over years to integrate new learning capabilities into important work settings.
This consortium, the MIT Center for Organizational Learning, now involves about
twenty corporations, mostly Fortune 100 firms.
Within the Learning Center companies, we regularly confront the dilemmas posed by
the conflicting views described above. The more we appreciate the inherent limitations
of executive leadership in bringing about deep change, the more frustrated we are likely
to feel, given the immense need for change. While top management can move some
changes quickly, it can also slow down or undermine other types of change. While
people often want the support of top management, they also dont want it telling them
what to do. Resolving these dilemmas requires fundamental shifts in our traditional
thinking about leadership. In brief, we are coming to believe that leaders are those
people who walk ahead, people who are genuinely committed to deep change in
themselves and in their organizations. They lead through developing new skills,
capabilities, and understandings. And they come from many places within an
organization.
In particular, we have come to think of three essential types of leaders in building
learning organizations, roughly corresponding to three different organizational
positions:

1. Local line leaders, who can undertake meaningful organizational experiments to test
whether new learning capabilities lead to improved business results
2. Executive leaders, who provide support for line leaders, develop learning
infrastructures, and lead by example in the gradual process of evolving the norms
and behaviors of a learning culture
3. Internal networkers, or community builders, the seed carriers of the new culture, who
can move freely about the organization to find those who are predisposed to
bringing about change, help out in organizational experiments, and aid in the
diffusion of new learnings

My purpose here is to briefly sketch what we are learning about these three types of
leaders. Previously (in a 1990 Sloan Management Review article), I proposed thinking
about the leaders new work in building learning organizations in terms of three

generic roles played by leaders at all levels, those of designer, teacher, and steward; the
skills and capabilities required for these roles; and the tools and methods that can help
in developing these skills and capabilities. Here, I want to begin to explore how these
generic roles are distributed among real people who lead from different organizational
positions.

Local Line Leaders

Nothing can start without committed local line leaders. Local line leaders are
individuals with significant business responsibility and bottom-line focus. They head
organizational units that are large enough to be meaningful microcosms of the larger
organization, and yet they have enough autonomy to be able to undertake meaningful
change independent of the larger organization. In effect, they create organizational
subcultures that may differ significantly from the mainstream organizational culture. To
be useful in creating experimental laboratories, they must also confront issues and
business challenges that are seen as both important and recurring within the larger
organization. For example, a unique cross-functional task force may be important but
less useful for a learning experiment than a team that manages a process that is ongoing,
generic, and vital for future competitiveness, such as a product development team, a
sales team, or a business division.
The key role played by local line leaders is to sanction significant practical
experiments and to lead through their active participation in those experiments. Without
serious practical experiments aimed at connecting new learning capabilities to business
results, there is no way to assess whether enhancing learning capabilities is just an
intellectually appealing idea or really makes a difference. Participating in serious
experiments requires a significant commitment of time and energy. Typically, a
Learning Center project will begin with a core team composed of line leaders who might
work together for six to twelve months developing their own understanding and skills
in systems thinking, collaborative inquiry into underlying mental models, and building
a shared vision, and applying those skills to their own issues. Only then will they be able
to begin to design learning processes that might spread such skills throughout their
organization and eventually become embedded in how work is done.
For example, a team of sales managers and sales representatives at Federal Express
worked together for over a year before they began to develop what eventually became
the Global Customer Learning Laboratory. We felt that we needed new tools for
working with our key corporate customers as learning partners, says Cathy Stopcynski
of Federal Express. Thats why the Global Customer Learning Laboratory is important.
It gives us a whole new way to work together with customers to improve our collective
thinking and come up with completely new solutions to complex logistics problems. At
Electronic Data Systems (EDS), a growing network of local line leaders is bringing

learning organization principles and methods into work with customers, through the
EDS Leading Learning Communities program.
In addition to playing a key role in the design and implementation of new learning
processes, local line leaders often become teachers once these learning processes become
established. We often find that the most effective facilitators in learning processes such
as the FedEx Learning Laboratory are not professional trainers but the line managers
themselves. Their substantive knowledge and practical experience give them unique
credibility. They are role models with whom other front-line people can identify. Last, as
the old saw goes, There is no better way to learn than to teach. Facilitating others
learning becomes a uniquely powerful way for line leaders to continually deepen their
own understanding and capabilities.
However, engaging local line leaders may be difficult. As pragmatists, they often find
ideas like systems thinking, mental models, and dialogue intangible and hard to get my
hands around. When I was first exposed to the MIT work, says Fred Simon, former
head of the Lincoln Continental program at Ford Motor Company, I was highly
skeptical. I had heard so many academic theories that made sense but never produced
for us. But I also was not happy with our teams ability to work together. I knew there
must be a better way, and my business planning manager was convinced this could
make a difference. Simons view is very typical of line leaders at the outset: he is
skeptical, but he recognizes that he has problems that he cannot solve, and he has a
trusted colleague who will engage with him. Again and again, we have found that
healthy, open-minded skeptics can become the most effective leaders and, eventually,
champions of this work. They keep the horse in front of the cart by focusing first and
foremost on business results. If they can find new approaches to enhance results, they
will commit time and energy to them. Such people invariably have more staying power
than the fans who get excited about new ideas but whose excitement wanes once the
newness wears off.
The limitations of local line leaders are natural counterparts to their strengths.
Because their focus is primarily on their business unit, they may not think much about
learning within the larger organization, and typically they have little time to devote to
diffusion of their efforts. They may also be unaware of and relatively inept at dealing
with antilearning forces in the larger organization. They can become impatient when the
larger organization does not change to match their new ideas of what works and may
start to feel misunderstood and unappreciated. They can easily develop an us against
the world siege mentality, which will then make them especially ineffective in
communicating their ideas to the unwashed.
Innovative local line managers are often more at risk than they realize. They typically
share a mental model that says, My bosses will leave me alone as long as I produce
results, regardless of the methods I use. But the better-mousetrap theory may not apply
in large institutions. Improved results are often threatening to others, and the more

dramatic the improvement, the greater the threat. Complex organizations have complex
forces that maintain the status quo and inhibit the spread of new ideas. Often, even the
most effective local line leaders fail to understand these forces or know how to work
with them.
Despite these limitations, committed local line leadership is essential. At least half of
the Learning Center companies that have made significant strides in improving business
results and developing internal learning capabilities and infrastructures have had little
or no executive leadership. In some instances, hundreds and literally thousands of
people are now involved in growing networks of internal practitioners with no active
top-management leadership. On the other hand, we have seen no examples where
significant progress has been made without leadership from local line managers, and
many examples where sincerely committed CEOs have failed to generate any significant
momentum.

Executive Leaders
Our fervor at the Learning Center with practical experiments led by local line managers
has frequently made us blind to the necessary complementary roles played by executive
leaders. We have gradually come to appreciate that local line leaders can benefit
significantly from executive champions who can be protectors, mentors, and thinking
partners. When dramatic improvements achieved in one line organization threaten
others, executive partners can help in managing the threat. Alternatively, new
innovative practices are often simply ignored because people are too busy to take the
time to really understand what the innovators are doing. Working in concert with
internal networkers, executives can help in connecting innovative local line leaders with
other like-minded people. They also play a mentoring role in helping the local line
leaders to mature, to understand complex political crosscurrents, and to communicate
their ideas and accomplishments to those who have not been involved.
For example, in one organization where a local line organization had achieved what
it regarded as dramatic improvements in the product development process, its process
improvement efforts lacked credibility outside the team when judged by more
traditional metrics. For instance, at critical checkpoints the team had record numbers of
engineering change orders. The team interpreted this as evidence that people were more
open, trusting, and willing to surface and fix problems early in the development process.
But outside the team, these same orders were seen as evidence of being out of control.
Eventually, executives in the company commissioned an independent audit, which
showed that the team was indeed highly effective. The executives also supported a
learning history to help others understand how the team had accomplished its results.
Part of the problem in appreciating effective executive leadership in learning is that
all of us are so used to the captain of the ship image of traditional hierarchical leaders.

We are used to thinking of top managers as the key decision makers, the most visible
and powerful people in the organization. Although undoubtedly some key decisions
will always have to be made at the top, cultures are not changed through singular
decisions, and decision-making power does not produce new learning capabilities.
When executives lead as teachers, stewards, and designers, they fill roles that are much
more subtle, contextual, and long term than the traditional model of the power-wielding
hierarchical leader suggests.
We in top management are responsible for the operating environment that can allow
continual learning, says Harley-Davidsons Teerlink. Although executive leadership
has traditionally focused on structure and strategy, Teerlink and other executives with
whom we at the Learning Center work are increasingly thinking about the operating
environment in less tangible ways. For example, over the past three years, Teerlink and
his colleagues have redesigned Harley-Davidsons traditional organizational structure
into three core circles of activity, each led by a leadership team. But according to
Teerlink, implementing the new organization was the easy part. Now we have to make
it work, and that requires people learning from each other.
Effective executive leaders build an operating environment for learning in several
ways. The first is through articulating guiding ideas. I have always believed that good
ideas will drive out bad ideas, said Hanovers Bill OBrien in a speech. One of the
basic problems with business today is that our organizations are guided by too many
mediocre ideas, ideas which do not foster aspirations worthy of peoples commitment.
Guiding ideas are different from slogans or the latest management buzzwords. They are
arrived at gradually, often over many years, through reflection on an organizations
history and traditions and on its long-term growth and opportunities.
The power of guiding ideas derives from the energy released when imagination and
aspiration come together. Understanding this power has always been a hallmark of
great leaders. The promise of learning organizations is, at least in part, the promise that
this power will become deeply and widely embedded in a way that rarely, if ever,
happens in traditional authoritarian organizations. There are two fundamentally
differing views of human nature and work, said OBrien. The objective view sees
work as a source of economic means. The subjective view is concerned with the effects
of work on the person. By the early twenty-first century, quality will become a
commodity and companies will be distinguished by the wholeness of their people.
A second way to build operating environments for learning is through conscious
attention to learning infrastructure. I believe that executives will increasingly come to
realize that, in a world of rapid change and increasing interdependence, learning is too
important to be left to chance. We have plenty of infrastructure for decision making
within AT&T, says Chairman Bob Allen. What we lack is infrastructure for learning
(Peter M. Senge, et al., The Fifth Discipline Handbook, 1994, p. 34).

I have met many CEOs in recent years who have lamented that we cant learn from
ourselves, that significant innovations simply dont spread, or that we are better at
learning from competitors than from our own people. Yet those very same executives
rarely recognize that they may be describing their own future job description. When we
stop to think, certain questions arise: Why should successful new practices spread in
organizations? Who studies these innovations to document why they worked? Where
are the learning processes whereby others might follow in the footsteps of successful
innovators? Who is responsible for these learning processes?
There can be little doubt of the long-term business impact of executive leadership in
developing learning infrastructure. When the Royal Dutch/Shell Groups central group
planning leaders became convinced that scenario thinking was a vital survival skill
in the turbulent, unpredictable world oil markets, they didnt initiate a set of scenarioplanning courses for Shells management. They redesigned the planning infrastructure
so that management teams regularly were asked not just for their budget and their
plan but for multiple plans describing how they would manage under multiple
possible futures. Planning as learning has gradually become a way of life within Shell,
a change to which many attribute Shells ascent to preeminence in the world oil
business.
A third way to build operating environments for business is the executives own
domain for taking actionnamely, the executive team itself. What is important, first, is
that executives see that they, too, must change, and that many of the skills that have
made them successful in the past can actively inhibit learning. They are forceful,
articulate advocates, but they usually are not very good at inquiring into their own
thinking or exposing the areas where their thinking is weak. Reflecting on a two-year
process both to rethink the corporate strategy and to develop new capacities to think
strategically together, Herman Miller CEO Kerm Campbell said, We started off as a
collection of individuals with strong views, which we often kept under wraps lest we
get into destructive battles with one another, and a set of assumptions we rarely
questioned about the keys to success in our industry. As we have progressed, I think we
have come to see that having an effective strategy and having the capacity to think
together strategically are inseparable.
How radical are ideas like these about executive leadership? I think they will
eventually lead to a very different mind-set and, ultimately, skill-set among executives.
Gradually, I have come to see a whole new model for my role as a CEO, says Shell
Oils Phil Carroll. Perhaps my real job is to be the ecologist for the organization. We must
learn how to see the company as a living system and to see it as a system within the
context of the larger systems of which it is a part. Only then will our vision reliably
include return for our shareholders, a productive environment for our employees, and a
social vision for the company as a whole.

Achieving such shifts in thinking, values, and behavior among executives is not easy.
The name of the game is giving up power, says Carroll. Even among enlightened
executives, giving up power is difficult. Its not that I dont miss the old system
sometimes, reflects Carroll. Being the commander in chief was kind of fun.

Internal Networkers
The most unappreciated leadership role is that of the internal networkers, or what we
often call internal community builders. Internal networkers are effective for the very
reasons that top-management efforts to initiate change can backfire. One of the most
interesting paradoxes in fostering deep change may be that no power is power.
Precisely because they have no positional authority, internal networkers are free to
move about a large organization relatively unnoticed. When the CEO visits someone,
everyone knows. When the CEO says, We need to become a learning organization,
everyone nods. But when someone with little or no positional authority begins inquiring
to see which people are genuinely interested in changing the way they and their teams
work, the only ones likely to respond are those who are genuinely interested. And if the
internal networker finds one person who is interested and asks, Who else do you think
really cares about these things, he or she is likely to receive an honest response. The
only authority possessed by internal networkers comes from the strength of their
convictions and the clarity of their ideas. This, we find time and again, is the only
legitimate authority when deep changes are required, regardless of ones organizational
position. The internal networkers have the paradoxical advantage that this is their only
source of authority.
It is very difficult to identify the internal networkers because they can be people from
many different organizational positions. They might be internal consultants, trainers, or
personnel staff in organization development or human resources. They might be frontline workers like engineers, sales representatives, or shop stewards. They might, under
some circumstances, be in senior staff positions. What is important is that they are able
to move around the organization freely, with high accessibility to many parts of the
organization. They understand the informal networks, what researchers call the
informal communities of practice, whereby information and stories flow and
innovative practices naturally diffuse within organizations (J. Seely Brown and P.
Duguid, Organization Science, Feb. 1991, pp. 4057).
Internal community builders work on different levels, commensurate with their
different levels in the organization. Front-line workers are typically the most effective
community builders within line organizations. Corporate staff people are often the most
effective networkers within larger corporations because they usually have a broad
perspective and extensive contacts. What matters is that effective internal networkers are
seen as credible, knowledgeable, committed individuals who are not a particular threat
to anyone. The most effective community organizer is he who is invisible, says Saul

Alinsky in Reveille for Radicals (1969), and the same can be said of the best internal
networkers.
The first vital function played by internal networkers is to identify local line
managers who have the power to take action and who are predisposed to developing
new learning capabilities. Much time and energy can be wasted working with the wrong
people, especially in the early stages of a change process. Convincing people that they
should be interested in systems thinking or learning is inherently a low-leverage
strategy. Even if they are persuaded initially, they are unlikely to persevere. When the
Liaison Officers from the Learning Center companies, a group that includes many of our
most effective internal networkers, asked, How did each of us get started in this work?
they responded, virtually unanimously, that they were predisposed. All of them had
something in their background perhaps an especially influential college course, a
particular work experience, or just a lifelong interestthat predisposed them to the
systems perspective, to a deep curiosity about learning or mental models or the mystery
of profound teamwork. In turn, they felt attuned to others they met who shared this
predisposition.
In ongoing experiments within line organizations, internal networkers can help in
many ways. In our own Learning Center projects, they serve as project managers, as
cofacilitators, or as learning historians, people trained to track a major change process
and to help those who are involved to better reflect on what they are learning. As
practical knowledge is built, internal networkers continue to serve as organizational
seed carriers, connecting people of like mind in diverse settings to each others
learning efforts. Gradually they may help in developing the more formal coordination
and steering mechanisms needed to leverage from local experiments to broader,
organizationwide learning. At Ford, for example, an informal Leaders of Learning group
was formed of multiple local line leaders to share learnings and to serve as a strategic
leadership body, supporting continuing experiments, connecting to the interests of top
management, and wrestling with the challenges of broader, organizationwide capacity
building and learning. Such groups, which gather together innovative line leaders, may
eventually evolve into an important element of the learning infrastructure. The initiative
to form this group came from Fords internal networkers.
The limitations of internal networkers likewise are not difficult to identify. Because
they do not have a great deal of formal authority, they can do little to directly counter
hierarchical authority. If a local line leader becomes a threat to peers or superiors, they
may be powerless to help her or him. Internal networkers have no authority to institute
changes in organizational structures or processes. So, even though they are essential,
they will be most effective in concert with local line leaders and executive leaders.

Conclusion

The leadership challenges in building learning organizations represent a microcosm of

the leadership issue of our times: how human communities, be they multinational
corporations or societies, productively confront complex, systemic issues where
hierarchical authority is inadequate for change. None of todays most pressing societal
issuesdeterioration of our natural environment, the international arms race, erosion of
the public education system, or the breakdown of the family and increasing social
anomie and fragmentationwill be resolved through hierarchical authority. In all these
issues, there are no simple causes, no simple fixes. There is no one villain to blame.
There will be no magic pill. Significant change will require imagination, perseverance,
dialogue, deep caring, and a willingness to change on the part of millions of people. I believe it is also the challenge posed in building learning organizations.
Recently a group of CEOs from the Learning Center companies spent a half-day with
Karl Henrik Robrt, the founder of Swedens path-breaking Natural Step process for
helping societies become ecologically sustainable. The next day, Rich Teerlink of HarleyDavidson came in and said, I dont know why I stay awake at night trying to figure out
how to transform a six-thousand-person company. Yesterday, we talked with someone
transforming a country of four million.
The challenges of systemic change where hierarchy is inadequate will, I believe, push
us to new views of leadership based on new principles. These challenges cannot be met
by isolated heroic leaders. They will require a unique mix of different people, in
different positions, who lead in different ways. Although the picture sketched above is
tentative and will undoubtedly evolve, I doubt that it understates the changes that will
be required in our traditional leadership models.

Note: This article is a shortened version of P. Senge, Leading Learning Organizations (MIT Center for
Organizational Learning Research Monograph), (Cambridge, Mass.: MIT Center for Organizational
Learning, 1995). Copyright 1995 by Peter M. Senge.